CEU Cardinal Herrera University,Centro de Estudios Universitarios

Chest X-ray (CXR) imaging is one of the most widely used diagnostic modalities in clinical practice, encompassing a broad spectrum of diagnostic tasks. Recent advancements have seen the extensive application of reasoning-based multimodal large language models (MLLMs) in medical imaging to enhance diagnostic efficiency and interpretability. However, existing multimodal models predominantly rely on ”one-time” diagnostic approaches, lacking verifiable supervision of the reasoning process. This leads to challenges in multi-task CXR diagnosis, including lengthy reasoning, sparse rewards, and frequent hallucinations. To address these issues, we propose CX-Mind, the first generative model to achieve interleaved ”think-answer” reasoning for CXR tasks, driven by curriculum-based reinforcement learning and verifiable process rewards (CuRL-VPR). Specifically, we constructed an instruction-tuning dataset, CX-Set, comprising 708,473 images and 2,619,148 samples, and generated 42,828 high-quality interleaved reasoning data points supervised by clinical reports. Optimization was conducted in two stages under the Group Relative Policy Optimization framework: initially stabilizing basic reasoning with closed-domain tasks, followed by transfer to open-domain diagnostics, incorporating rule-based conditional process rewards to bypass the need for pretrained reward models. Extensive experimental results demonstrate that CX-Mind significantly outperforms existing medical and general-domain MLLMs in visual understanding, text generation, and spatiotemporal alignment, achieving an average performance improvement of 25.1% over comparable CXR-specific models. On a real-world clinical dataset (Rui-CXR), CX-Mind achieves a mean recall@1 across 14 diseases that substantially surpasses the second-best results, with multi-center expert evaluations further confirming its clinical utility across multiple dimensions. CX-Mind establishes a new paradigm for constructing interpretable, and high-performing medical MLLMs.

The aim of this study was to analyse and compare the biomechanical performance of a novel drilling technique known as osseodensification, which focuses on bone preservation and compaction rather than removal, against the conventional osteotomy technique in implant placement. Online databases were searched to identify relevant articles published in English during the last 10 years (up until September 11, 2025), reporting in vivo clinical trials involving a healthy adult population undergoing single or multiple implant placement, and comparing osseodensification and standard osteotomy drills, in a partially or fully edentulous segment of alveolar bone. Biomechanical measures analysed were the insertion torque and implant stability quotient (ISQ). Risk of bias and study quality were evaluated using the RoB 2 tool, Newcastle-Ottawa Scale, and GRADE system. The results were synthesized. Overall, a total of 555 patients (59.2% female, in studies reporting the sex distribution), received a total of 685 implants (343 using osseodensification). The overall mean insertion torque in the osseodensification group was 45.75 ± 6.55 N·cm, while in the standard osteotomy group it was 38.00 ± 7.01 N·cm (P < 0.001). The estimated average mean difference in ISQ (random-effects model) was 3.24 (95% confidence interval 0.72-5.95; P = 0.024). The data showed that sites prepared with osseodensification resulted in higher implant torque and primary stability of the implant (ISQ) immediately postoperative. However, longitudinal data were insufficient to allow a meaningful statistical analysis of ISQ trends over time.

BackgroundOpen wounds are a common clinical problem in small animal practice and are associated with delayed healing, infection, and increased treatment burden. Fluorescence biomodulation (FB), a form of photobiomodulation therapy (PBMT), has shown benefits in several dermatological conditions; however, its efficacy in veterinary open wounds remains unclear.ObjectivesTo evaluate the effect of FB as an adjunct to standardised wound management on healing progression in open wounds in dogs and cats using a prospective, randomised, within-wound controlled clinical trial.MethodsFifteen client-owned dogs and cats with open wounds of various aetiologies were enrolled. Each wound was divided into two portions: one treated weekly with FB and one untreated control, both receiving standardised wound care based on the TIME framework. Quantitative wound area reduction was assessed by serial planimetric measurements. Qualitative healing parameters were scored weekly by two blinded observers, and semiquantitative histological and immunohistochemical evaluation was performed in surgically treated cases. Paired statistical analyses were used. A time-integrated analysis based on the area under the curve (AUC) of normalised wound area was performed to assess healing kinetics.ResultsNo statistically significant differences were observed between treated and control portions for percentage wound area reduction, qualitative healing scores, or semiquantitative histological and immunohistochemical outcomes. Quantitative planimetry showed a heterogeneous response, with seven wounds favouring the treated portion, six favouring the control portion, and two showing equivalent outcomes. Exploratory post hoc time-integrated analyses suggested a trend toward earlier wound area reduction in some cases, particularly when treatment was initiated early; however, these findings were not statistically significant. Complete wound closure during follow-up did not differ significantly between portions. Histological analysis showed only small numerical differences between portions, with no statistically significant differences in collagen deposition, angiogenesis, inflammatory response, fibroblast proliferation, or CD31 and Ki67 expression.ConclusionUnder the conditions tested, FB did not demonstrate a consistent measurable advantage over standardised wound management in open wounds in dogs and cats. Any potential clinical benefit appeared modest and possibly limited to selected wound presentations. Further prospectively stratified studies focused on healing speed and readiness for closure are warranted.

BACKGROUND:The introduction of smallpox vaccination in the early nineteenth century marked a major milestone in the history of public health in Spain. Although its early dissemination has been mainly studied from an institutional perspective, the contribution of local physicians to its implementation remains less well understood. METHODS:Historical study based on documentary sources conducted using primary sources, including archival records, historical newspapers, and medical memoirs, together with relevant secondary literature. The study focuses on the career of the Navarrese physician Manuel Gil y Albéniz (1767-1849), who practised in the town of Cascante (Navarre). RESULTS:Gil y Albéniz introduced smallpox vaccination in Cascante in 1801, during the early stage of its adoption in Spain. He carried out hundreds of vaccinations and played an important role in disseminating vaccine material throughout Navarre. His writings describe methods for preserving and transmitting vaccine material and document strategies that enabled vaccination to continue during the Peninsular War. He also participated in scientific networks as a corresponding member of the Royal Academy of Medicine of Barcelona (1803) and the Medical-Surgical Society of Cádiz (1820). His Colección de memorias médicas (1820) brings together his principal clinical observations. CONCLUSIONS:Manuel Gil y Albéniz was among the earliest physicians to introduce smallpox vaccination in Navarre. His career highlights the important contribution of local physicians in the dissemination of vaccination and the early development of public health in nineteenth-century Spain.